Last updated: 2022-01-11

Checks: 6 1

Knit directory: DepInfeR/analysis/

This reproducible R Markdown analysis was created with workflowr (version 1.7.0). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.

R Markdown file: uncommitted changes

The R Markdown is untracked by Git. To know which version of the R Markdown file created these results, you’ll want to first commit it to the Git repo. If you’re still working on the analysis, you can ignore this warning. When you’re finished, you can run wflow_publish to commit the R Markdown file and build the HTML.

Environment: empty

Great job! The global environment was empty. Objects defined in the global environment can affect the analysis in your R Markdown file in unknown ways. For reproduciblity it’s best to always run the code in an empty environment.

Seed: set.seed(20211005)

The command set.seed(20211005) was run prior to running the code in the R Markdown file. Setting a seed ensures that any results that rely on randomness, e.g. subsampling or permutations, are reproducible.

Session information: recorded

Great job! Recording the operating system, R version, and package versions is critical for reproducibility.

Cache: none

Nice! There were no cached chunks for this analysis, so you can be confident that you successfully produced the results during this run.

File paths: relative

Great job! Using relative paths to the files within your workflowr project makes it easier to run your code on other machines.

Repository version: 43be8a7

Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility.

The results in this page were generated with repository version 43be8a7. See the Past versions tab to see a history of the changes made to the R Markdown and HTML files.

Note that you need to be careful to ensure that all relevant files for the analysis have been committed to Git prior to generating the results (you can use wflow_publish or wflow_git_commit). workflowr only checks the R Markdown file, but you know if there are other scripts or data files that it depends on. Below is the status of the Git repository when the results were generated: 


Ignored files:
    Ignored:    .DS_Store
    Ignored:    .Rhistory
    Ignored:    .Rproj.user/
    Ignored:    analysis/.DS_Store
    Ignored:    analysis/.Rhistory
    Ignored:    analysis/analysis_RNAseq_cache/
    Ignored:    data/.DS_Store
    Ignored:    output/.DS_Store

Untracked files:
    Untracked:  analysis/analysis_EMBL2016.Rmd
    Untracked:  analysis/analysis_GDSC.Rmd
    Untracked:  analysis/analysis_RNAseq.Rmd
    Untracked:  analysis/analysis_beatAML.Rmd
    Untracked:  analysis/process_EMBL2016.Rmd
    Untracked:  analysis/process_GDSC.Rmd
    Untracked:  analysis/process_beatAML.Rmd
    Untracked:  analysis/process_kinobeads.Rmd
    Untracked:  code/utils.R
    Untracked:  data/BeatAML/
    Untracked:  data/EMBL2016/
    Untracked:  data/GDSC/
    Untracked:  data/Kinobeads/
    Untracked:  data/RNAseq/
    Untracked:  manuscript/
    Untracked:  output/BeatAML_result.RData
    Untracked:  output/EMBL_result.RData
    Untracked:  output/EMBL_resultSub.RData
    Untracked:  output/GDSC_result.RData
    Untracked:  output/allTargets.rds
    Untracked:  output/inputs_BeatAML.RData
    Untracked:  output/inputs_EMBL.RData
    Untracked:  output/inputs_GDSC.RData

Unstaged changes:
    Modified:   README.md
    Modified:   _workflowr.yml
    Modified:   analysis/_site.yml
    Deleted:    analysis/about.Rmd
    Modified:   analysis/index.Rmd
    Deleted:    analysis/license.Rmd
    Deleted:    output/README.md

Note that any generated files, e.g. HTML, png, CSS, etc., are not included in this status report because it is ok for generated content to have uncommitted changes.

There are no past versions. Publish this analysis with wflow_publish() to start tracking its development.

Load packages

Packages

library(DepInfeR)
library(RColorBrewer)
library(pheatmap)
library(ggbeeswarm)
library(ggrepel)
library(fpc)
library(igraph)
library(factoextra)
library(tidyverse)
source("../code/utils.R")
knitr::opts_chunk$set(dev=c("png","pdf"))

Load pre-processed datasets

load("../output/inputs_EMBL.RData")

Dimensions of input matrices

Drug-target

dim(tarMat_EMBL)
[1]  85 136

Drug-sample (viability matrix)

dim(viabMat_EMBL)
[1]  85 131

Multivariant model for protein dependence prediction

Perform multivariant LASSO regression based on a drug-protein affinity matrix and a drug response matrix.
This chunk can take a long time to run. Therefore we will save the result for later use to save time.

set.seed(333)
#column wise scale of the viability matrix, to keep the drug effect rank
viabMat.scale <- t(mscale(t(viabMat_EMBL)))

#run lasso regression
result <- runLASSOregression(TargetMatrix = tarMat_EMBL , ResponseMatrix = viabMat.scale)

#remove targets that were never selected
useTar <- rowSums(result$coefMat) != 0
result$coefMat <- result$coefMat[useTar,]


#save intermediate results
save(result, file = "../output/EMBL_result.RData")

Load the saved result

load("../output/EMBL_result.RData")

Number of selected targets

nrow(result$coefMat)
[1] 24

Make genomic patient annotation

Prepare column annotations

#genetic background annotation
colAnno <- dplyr::select(annotation_EMBL, Patient.ID, 
                  diagnosis, IGHV.status, 
                  Methylation_Cluster,
                  trisomy12, TP53, del11q) %>% 
  data.frame() %>% remove_rownames() %>%
  column_to_rownames("Patient.ID")

colAnno_cll <- dplyr::filter(colAnno, diagnosis == "CLL")


#color for annotation
annoColor <- list(treatment = c(Yes = "black", No = "grey80"),
                  IGHV.status = c(M = "black",U="grey80"),
                  Methylation_Cluster = c(HP = "darkblue", IP = "blue", LP = "lightblue"),
                  diagnosis = c(CLL = "#BC3C29FF", MCL = "#E18727FF",`T-PLL`="#20854EFF"))
for (name in setdiff(colnames(colAnno),names(annoColor))) {
  annoColor[[name]] <- c(`1` = "black",`0` = "grey80") 
}

Heatmap plot of protein dependence matrix

plotTab <- result$coefMat

#normalization for different protein dependencies (over samples) without changing the coefficient sign
plotTab_scaled <- scale(t(plotTab), center = FALSE, scale = TRUE)
plotTab <- t(plotTab_scaled)

pheatmap(plotTab, 
         color=colorRampPalette(rev(brewer.pal(n = 7, name ="RdBu")), bias= 1.2)(100),
         annotation_col = colAnno,
         annotation_colors = annoColor,
         clustering_method = "ward.D2", scale = "none", 
         show_colnames = TRUE, main = "row scaled", fontsize = 9, fontsize_row = 10, fontsize_col = 7)

BTK importance among CLL, MCL and T-PLL

coefMatAll <- result$coefMat
plotTab <- tibble(patID = colnames(coefMatAll),
                  coef = coefMatAll["BTK",]) %>%
  mutate(diagnosis = colAnno[patID,]$diagnosis) %>%
  group_by(diagnosis) %>% mutate(n=length(patID)) %>%
  mutate(diagnosis = sprintf("%s (n=%s)", diagnosis, n))

g <- ggplot(na.omit(plotTab), aes(x=diagnosis,y=coef)) + 
  geom_violin(aes(fill = diagnosis)) + scale_fill_manual(values=c("#46B8DAFF","#D43F3AFF","#EEA236FF")) + geom_beeswarm() +
  xlab("") + ylab("Protein dependence") + ggtitle(sprintf("BTK ~ diagnosis")) +
  theme_custom + theme(legend.position = "none")



g

Assessment of results

PCA

PCA plot of protein dependence matrix of CLL samples colored by IGHV status

coefMat_cll <- result$coefMat[,rownames(colAnno_cll)]

res_pca_imp_kd_cll <- prcomp(t(coefMat_cll), scale = T, center = TRUE)
#fviz_eig(res_pca_imp_kd_cll)

#PCA plot for IGHV mutation status
fviz_pca_ind(res_pca_imp_kd_cll,
             geom = c("point"),
             pointsize = 2,
             repel = TRUE,  
             xlab = "PC1",
             ylab = "PC2", 
             habillage = colAnno_cll$IGHV.status,
             title = "Protein dependence - PCA - Lasso kd - CLL") +
                  theme(axis.text = element_text(size = 12), 
                        axis.title = element_text(size = 14),
                        plot.title = element_text(size = 12, hjust = 0.5, face = "bold")) + 
                  scale_color_manual(breaks=c("M","U"), values= c("#46B8DAFF","#D43F3AFF"), name = "IGHV") +  
                  scale_shape(guide=FALSE)
Warning: Removed 2 rows containing missing values (geom_point).
Warning: Removed 1 rows containing missing values (geom_point).
Warning: It is deprecated to specify `guide = FALSE` to remove a guide. Please
use `guide = "none"` instead.

k-means clustering

k-means clustering of protein dependence matrix with IGHV status annotation

colList_mut <- c("#BC3C29FF","#0072B5FF")
kclus_tab <- merge(colAnno_cll[,1:2], t(coefMat_cll), all = T, by = 'row.names')
kclus_tab <- remove_rownames(kclus_tab) %>% column_to_rownames("Row.names")
kclus_tab <- na.omit(kclus_tab)
km_res_imp <- eclust(kclus_tab[,-c(1:2)], "kmeans", k = 2, nstart = 50, graph = FALSE, stand = T)

pcTar <- fviz_cluster(km_res_imp, kclus_tab[,-c(1:2)], geom = c("point")) + 
  geom_point(aes(colour= kclus_tab$IGHV.status)) + 
  ggtitle("k-means clustering of \nprotein dependence matrix") +
  scale_shape(guide=FALSE) +   
  scale_color_manual(breaks=c("M","U"), values= c("#0072B5FF", "#BC3C29FF", "#BC3C29FF","#0072B5FF"), name = "IGHV") +
  scale_fill_manual(values = colList_mut) + xlab("PC1") + ylab("PC2") +
  theme_custom + theme(plot.title = element_text(size=20, face="bold")) + xlim(-6,6) + ylim(-8,8)

pcTar
Warning: It is deprecated to specify `guide = FALSE` to remove a guide. Please
use `guide = "none"` instead.

k-means clustering of drug response matrix with IGHV status annotation

kclus_tab_drug <- merge(colAnno_cll[,1:2], t(viabMat_EMBL), all=T, by='row.names')
kclus_tab_drug <- remove_rownames(kclus_tab_drug) %>% column_to_rownames("Row.names")
kclus_tab_drug <- na.omit(kclus_tab_drug)

km_res_drug <- eclust(kclus_tab_drug[,-c(1:2)], "kmeans", k = 2, nstart = 50, graph = FALSE, stand = T)

pcDrug <- fviz_cluster(km_res_drug, kclus_tab_drug[,-c(1:2)], geom = c("point")) + 
  geom_point(aes(colour= kclus_tab_drug$IGHV.status)) + 
  ggtitle("k-means clustering of \ndrug response matrix") +
  scale_shape(guide=FALSE) +   scale_color_manual(breaks=c("M","U"), values= c("#0072B5FF", "#BC3C29FF", "#BC3C29FF","#0072B5FF"), name = "IGHV") +
  scale_fill_manual(values = colList_mut) + xlab("PC1") + ylab("PC2") +
  theme_custom + theme(plot.title = element_text(size=20, face="bold")) +
  scale_x_reverse(breaks = c(12, 6, 0, -6, -12), limits=c(12,-12), labels = c(-12,-6,0,6,12)) + #mirro x-axis for the sake of visualization
  ylim(-12,12)
pcDrug
Warning: Removed 4 rows containing non-finite values (stat_chull).
Warning: Removed 4 rows containing non-finite values (stat_mean).
Warning: Removed 4 rows containing missing values (geom_point).

Warning: Removed 4 rows containing missing values (geom_point).
Warning: It is deprecated to specify `guide = FALSE` to remove a guide. Please
use `guide = "none"` instead.

External cluster assessment (based on k-means results) - IGHV status - Rand Index

An external cluster analysis evaluates the clustering quality by comparing the found cluster belonging to the known ground-truth. The Rand Index is a good measure to evaluate how many datapoints were clustered correctly when comparing the found clusters to the known ground truth. The Rand Index ranges between 0 and 1 and a higher index corresponds to more datapoints being clustered in their correct groups.

# function to calculate cross tabulation between k-means clusters and ground truth
externalClust <- function(clusterTab, kmeansResult, header) {
  #calculation
  crossTab <- table(clusterTab$IGHV.status, kmeansResult$cluster)
  status <- as.numeric(factor(clusterTab$IGHV.status))
  imp_stats <- cluster.stats(d = dist(clusterTab[,-c(1:2)]), 
                             status, kmeansResult$cluster)
  randScore <- round(imp_stats$corrected.rand, 3)
  #create table
  crossTab <- data.frame(IGHV = rownames(crossTab), Cluster1 = c(crossTab[,1]), Cluster2 = c(crossTab[,2])) %>% 
    gt::gt(rowname_col = "rowname") %>%  
    gt::tab_header(title = header, subtitle = paste("Rand Index is", randScore)) 
  return(crossTab)
}


# target importance
externalClust(kclus_tab, km_res_imp , header = "Cross tabulation protein dependence")
Cross tabulation protein dependence
Rand Index is 0.623
IGHV Cluster1 Cluster2
M 55 5
U 7 48
 
# drug screen
externalClust(kclus_tab_drug, km_res_drug , header = "Cross tabulation drug screen")
Cross tabulation drug screen
Rand Index is 0.146
IGHV Cluster1 Cluster2
M 47 13
U 22 33

Differential dependence on proteins associated with genotype

Prepare genomic background table

patAnno <- dplyr::select(annotation_EMBL, -sex, -treatment, -date.of.first.treatment)

Association test between protein dependence and genetic background

geneBack <- patAnno %>% data.frame() %>% remove_rownames() %>%
    column_to_rownames("Patient.ID")

testRes_cll <- diffImportance(coefMat_cll, geneBack)

Visualization of identified associations

#define color
colList <- c(`not significant` = "grey80", "Tris12 positive / U-IGHV" = "#BC3C29FF", "Tris12 negative / M-IGHV" = "#0072B5FF")
pos = position_jitter(width = 0.25, seed = 10)

plotTab <- testRes_cll %>% dplyr::filter(mutName %in% c("IGHV.status", "trisomy12")) %>% mutate(type = ifelse(p.adj > 0.1, "not significant", 
                                          ifelse(FC >0, "Tris12 positive / U-IGHV","Tris12 negative / M-IGHV"))) %>%
  mutate(varName = ifelse(type == "not significant","",targetName)) 

plotTab <- plotTab %>% dplyr::filter(mutName %in% plotTab$mutName[1:19]) %>%
  mutate(mutName = str_replace(mutName,"[.]"," "))

p <- ggplot(data=plotTab, aes(x= mutName, y=-log10(p.adj),
                              col=type, label = varName))+ 
  geom_hline(yintercept = -log10(0.1), linetype="dotted", color = "grey20") + 
  geom_point(size=3, position = pos) + 
  geom_text_repel(position = pos, color = "black", size= 6, force = 3) +
  ylab(expression(-log[10]*'('*adjusted~italic("P")~value*')')) + xlab("") +
  scale_color_manual(values = colList) +
  theme_custom +
  #annotate(geom = "text", x = 0.5, y = -log10(0.1) - 0.25, label = "10% FDR", size=7, col = "grey20") +
  coord_flip()  + labs(col = "Higher dependence in")  +
  theme(legend.position = c(0.75,0.8),
        legend.background = element_rect(fill = NA),
        legend.text = element_text(size=14),
        legend.title = element_text(size=16),
        axis.title = element_text(size=18),
        axis.text = element_text(size=18))
plot(p)
Warning: ggrepel: 2 unlabeled data points (too many overlaps). Consider
increasing max.overlaps

#ggsave("test.pdf", height = 4, width = 8)

Visualize significant associations using a heatmap

plotTab <- testRes_cll %>% dplyr::filter(mutName %in% c("IGHV.status", "trisomy12")) %>% 
  mutate(starSign = ifelse(p.adj <=0.1, "*", ""),
         pSign = -log10(p)*sign(FC))

#subset for mutation with at least one significant associations
plotTar <- unique(filter(plotTab, p.adj <= 0.1)$targetName)
plotMut <- unique(filter(plotTab, p.adj <= 0.1)$mutName)
plotTab <- plotTab %>% dplyr::filter( targetName %in% plotTar , mutName %in% plotMut) 


p <- ggplot(data=plotTab, aes(y=mutName, x = targetName, fill=pSign)) +
  geom_tile(col = "black") + geom_text(aes(label = starSign), size=5, vjust=0.5) +
  scale_fill_gradient2(low = "#BC3C29FF", high = "#0072B5FF", name = bquote(-log[10]*italic("P"))) +
  theme_minimal() +
  theme(panel.grid.major = element_blank(),
        legend.text = element_text(size=14),
        legend.title = element_text(size=16),
        axis.title = element_text(size=18),
        axis.text.y = element_text(size=18),
        axis.text.x = element_text(size=18, angle = 90, vjust=.5, hjust=1)) +
  ylab("Mutations") + xlab("Proteins")

p

Visualization of exemplary associations in beeswarm plots

pList <- plotDiffBox(testRes_cll, coefMat_cll, geneBack, fdrCut = 0.05)
cowplot::plot_grid(pList$BTK_IGHV.status, 
                   pList$SIK2_IGHV.status + theme(axis.title.y = element_blank()), 
                   pList$INPPL1_IGHV.status + theme(axis.title.y = element_blank()), 
                   pList$MAP2K2_trisomy12+theme(axis.title.y = element_blank()), nrow=1,
                   rel_widths = c(1.05,1,1,1))